The need for continuous duty, high power rotating anode x-ray tubes exists in medical radiography, i.e, fluoroscopy and computerized tomography (CT), and in industrial applications such as x-ray diffraction topography and non-destructive testing.
A number of schemes have been proposed in the past to achieve continuous power output at high peak power with a rotating anode x-ray tube. These include direct liquid cooling of the anode, liquid to vapor phase cooling of the anode, as well as other techniques.
A prior art scheme for liquid cooling rotating anodes is described in the Philips Technical Review, Vol. 19, 1957/58, No. 11, pp. 314-317. The rotating anode of the Philips device constitutes a hollow cylinder with three radially running tubes through which water flows to a cavity located along the inner surface of the peripheral wall or anode strip of the hollow body. In this device, the water flows back into the hollow drive shaft through three other tubes running radially in the rotary anode. However, various disadvantages have been attributed to the Philips device. For example, U.S. Pat. No. 4,130,772 to Kussel, et al., issued December 1978, states that only relatively low speeds of rotation can be obtained with the Philips rotary device because the maximum thickness of the peripheral wall provided as the anode target member allowable for proper cooling is not sufficient to withstand the pressures in the cooling medium that arise due to centrifugal force at higher speeds of revolution. Only relatively small surface density of illumination (brightness) can be obtained with this known rotary anode, since the intensity of illumination, i.e., radiation per unit of surface, generated by a device depends upon the rate of anode revolution.
The Kussel, et al. patent describes a liquid cooled rotating anode which purports to resolve the shortcomings of the Philips device. The portion of the rotary anode cylindrical peripheral wall, whereon the electron beam strikes, is cooled with water supplied and removed, respectively, through coaxial ducts distributed by radial ducts in one end face of the rotary to a ring duct and gathered from a ring duct as the other end face through another set of radial ducts leading back to the shaft. Between the two ring ducts, the cooling medium flows through helical cooling ducts running parallel to each other and at an angle of about 15.degree. to the edge boundaries of the cylindrical operating surface. These ducts are formed on the outside by the anode peripheral wall material itself and on the inside by a stainless steel insert.
The Kussel device, although resolving the shortcomings of the Philips device, has several problems of its own--one of them, basic. To obtain efficient heat transfer, relatively high coolant velocities are required. To achieve high coolant velocities, high pump pressures are needed. Unfortunately, the seals necessary to join stationary to rotating fluid conduits generally have short lives when subjected to such high coolant pressures and high speed anode rotation.
A more basic limitation of the Kussel et al. device arises from the use of the metal insert with grooves machined thereon to form the coolant ducts. The outermost rims of the groove walls are brazed to the anode peripheral wall. As described, the cooling ducts traverse one face of the anode to the other at a pitch angle of 15.degree.. Therefore, the duct walls whose peripheries are brazed to the inside surface of the anode opposite the electron beam track also traverse one face of the anode to the other at the prescribed 15.degree. angle. Therefore, the electron beam alternately travels over coolant duct and then duct wall as the anode rotates. When the electron beam is above the coolant, heat transfer is efficient, whereas when it is above the duct wall, it simulates more closely a solid metal structure, i.e., a conventional solid rotating anode. This creates a hot spot and severely limits the power handling capability because of the long heat path to the coolant. The braze alloy, used to braze the anode to the insert and which must melt well below the metals used, further limits the power densities that can be handled. The duct walls, brazed to the periphery of the anode, which provide the necessary strength to the anode shell to prevent its distortion due to centrifugal force of the coolant, become a liability in that they become a limiting factor in power handling capability.
U.S. Pat. No. 4,165,472, issued on Aug. 21, 1979, to Wittry describes a device utilizing a cooling technique typically referred to as "liquid to vapor phase cooling." In the preferred embodiment of the Wittry patent, a two-stage system is used. The first stage consists of a sealed chamber in the anode that is filled with a coolant, such as water, that removes heat by vaporizing and recondensing on another portion of the internal anode surface that is cooled by a secondary liquid cooling loop. This in turn removes the heat to a heat sink external to the x-ray tube. In general, the various embodiments described are described as wickless heat pipes. One limitation is that heat transfer is limited by the diffusion rate of the vapor phase to the cool surface. A 6 kw capability is described in terms of a 12" diameter anode rotating at 5000 rpm. Directly cooled rotating anode x-ray tubes are rated at higher powers. Kussel discloses power capability of 100 kw. A further limitation on this structure is the sealed coolant chamber. A small amount of overheating can cause excessive pressures to be built up, i.e., bearing wear slowing the rotation. If the structure does not explode, it will bulge which will throw it out of balance, thereby rapidly wearing out the bearings.
U.S. Pat. No. 3,959,685, issued on May 25, 1976 to Konieczynski discloses a method whereby the heat capacity of a conventional, solid rotating anode x-ray tube can be increased. This is accomplished by sealing slugs of high heat capacity and selected melting point metal into the anode. When the anode reaches a critical temperature, the slugs melt, absorbing more heat. Upon cooling, they re-solidify. A 20% increase in heat capacity is mentioned. The limitation of this device is that should the melted slugs overheat and create excessive pressures due to target slowdown or stoppage (frozen bearings), it truly becomes a bomb with molten metal spewing out. This makes it unacceptable for medical use. Any irregularities in resolidification of the slugs, due to small differences in cooling rates or irregular crystal formation, will cause an imbalance in the anode with resultant early bearing failure.
U.S. Pat. No. 3,719,847 issued on Mar. 6, 1973 to Webster provides a hollow anode in which a liquid metal such as sodium or lithium is confined. Heat from the electron beam is striking the cathode which causes the liquid metal to evaporate, thereby effectively increasing the heat capacity of the anode. With no means to extract the heat, cooling is by radiation as with a conventional solid anode. Should the anode overheat, due to bearing wear, etc., the confined metal vapor will build up excessive pressure and the vessel can explode with consequent danger to personnel in the vicinity.
U.S. Pat. No. 4,146,815, issued Mar. 27, 1979, to Childenc, also discloses a hollow anode filled with a liquid metal much like that disclosed in Webster. It suffers from the same limitation of retaining the characteristic of a solid anode that must cool by radiation. It also possesses the potential of exploding like a bomb should it overheat due to bearing wear caused by age or imbalance.
U.S. Pat. No. 3,735,175, issued May 22, 1973, to Blomgren, discloses a heat pipe to transmit heat from the anode to an external heat sink. Notwithstanding the efficacy of external electrostatic cooling, a heat pipe depends on the diffusion rate of the coolant vapor to the cool end for the rate of heat removal. The power densities that can be handled are relatively low. For the power levels required, a huge and impractical heat pipe would be needed, i.e., 50 kw dissipation.
U.S. Pat. No. 3,794,872, issued Feb. 26, 1974 to Haas, discloses a fixed target anode cooled by a jet of fluid. The target is mounted on a bellows such that "the target reciprocates laterally in a direction perpendicular to the axis of the tube but the target does not rotate on its own axis." As the focal spot wears out, i.e., pits, the target is moved to a new position to provide fresh target surface. In this manner, the effective life of the tube is extended considerably. The motion provided is not rotational and therefore does not increase the output power of the tube. As a fixed target tube, its power output is low.
A prior art alternative to the respective Philips and Kussel et al. approaches to dissipation of large power loads is that of Taylor as described in Advances in X-ray Analysis, Vol. 9, August 1965, G. R. Mallett, et al., Plenum Press, N.Y. In the Taylor design, the liquid coolant flows transverse to the direction of anode rotation and interacts with the anode in a manner known as "linear coolant flow." However, although there is a high relative velocity between the anode and coolant, the interaction is relatively inefficient and is reported by Taylor to provide only relatively low power (71/2 kw). This stands in sharp contrast to the 100 kw attributed to the Kussel design. However, the Taylor design is not subject to performance-limiting centrifugal forces as the Philips device is, and permits the use of low pressure pumps and components.
Further description of prior art liquid cooled rotating anode x-ray tubes is found in the following articles:
G. Fournier, J. Mathieu: J. Phys. 8, 177 (1937) PA0 R. E. Clay: Proc. Phys. Soc. (London) 46, 703 (1934) PA0 R. E. Clay, A. Miller: J. Instr. Elect. Engs. 84, 261 (1939) PA0 W. T. Astbury, R. D. Preston: Nature 24, 460 (1934) PA0 Z. Nishiyama: J. Japan Met. Soc. 15,42 (1940) PA0 V. Linnitzki, V. Gorski: Sov. Phys-Tech, Phys. 3, 220 (1936) PA0 R. R. Wilson: Rev. Sci. Instr. 12,91 (1941) PA0 S. Miyake, S. Hoshino: X-sen 8,45 (1954) (Japanese) PA0 Y. Yoneda, K. Kohra, T. Futagami, M. Koga: Kyushu Univ. Engs. Dept. Rep. 27,87 (1954) PA0 S. Kiyono, M. Kanayama, T. Konno, N. Nagashita: Technol. Rep., Tohoku Univ. XXVII, 103 (1936) PA0 A. Taylor: J. Sci. Instr. 26,225 (1949); Rev. Sci. Instr. 27,757 (1956) PA0 D. A. Davies: Rev. Sci. Instr. 30,488 (1959) PA0 P. Gay, P. B. Hirsh, J. S. Thorp, J. N. Keller: Proc. Phys. Soc. (London) B64,374 (1951) PA0 A. Muller: Proc. Roy. Soc. 117,31 (1927) PA0 W. T. Astbury: Brit. J. Rad. 22,360 (1949) PA0 E. A. Owen: J. Sci. Instr. 30,393 (1953) PA0 K. J. Queisser: X-ray Optics, Applications to Solids Verlag-Springer, NY (1977), Chap. 2 PA0 Longley: Rev. Sci. Instr. 46,1 (1975) PA0 Mayden: Conference on Microlithography, Paris, June 21-24, 1977, pp. 196-199 PA0 MacArthur: Electronics Eng. 17,1 (1944-5) PA0 A. E. DeBarr: Brit. J. Appl. Phys. 1,305 (1950)